Cigarette smoking continues to be a serious public health issue that puts a considerable portion of adults in the US at risk for life-threatening diseases. The vast majority of smokers initiate smoking before age 16, underscoring the need to prospectively study the development of smoking behavior in adolescents. Ongoing research has identified several consistent risk factors of adolescent smoking; including parental smoking, peer smoking, depression, and nicotine dependence (particularly early-emerging nicotine dependence). There is ample evidence that many of these risk factors influence each other over time and have time-varying effects on smoking outcomes, yet these relationships cannot be investigated using standard analyses. Thus, a major gap in current knowledge about smoking initiation and progression is the absence of interrelationships among risk factors in most analyses, which can affect the course of smoking behavior. Further, causal relationships cannot be tested using standard regression analyses of observational data. Systems methods, such as system dynamics modeling, are capable of addressing these questions by simulating smoking behavior over time as a system of interconnected, dynamic variables which contains causal links and feedback loops. This proposal contains both a training component and a substantive component which will investigate the development of adolescent smoking behavior. For the training component, the applicant will augment her introductory training in system dynamics modeling by taking a number of formal courses. Further, the applicant will meet throughout the award period with a team of consultants, which consists of experts in traditional system dynamics modeling, the wider family of systems methods, or the field of adolescent smoking. Consulting meetings will bridge the formal training with the substantive component, in that the consultants will supervise the models developed by the applicant and the applicant's interpretation of results. For the substantive component of this proposal, a comprehensive model of smoking development and its major risk factors will be developed incrementally over the Specific Aims. For Aim 1, the positive feedback loop between nicotine dependence and smoking behavior, and its subsequent effect on progression from experimental smoking to regular smoking, will be investigated. This model will be expanded in Aim 2 to include parental smoking's effect on initiation and progression rates and nicotine dependence. Aim 2 has additional smoking statuses (nonsmokers, former smokers) and corresponding transitions (smoking initiation, cessation of experimentation, quitting regular smoking). Aim 3 will investigate the sensitivity of the system to the initial level of nicotine dependence; this will be modeled by adding depression and peer smoking to the model developed in Aim 2. This project will significantly contribute to current understanding of the development of adolescent by elucidating the complex and time-varying interrelationships among risk factors. This project could provide valuable recommendations to public health policy and/or clinical settings by identifying which strategies are likely to be effective targets of prevention or intervention strategies, and the optimal time to employ them. Finally, an accurate, calibrated model of smoking development can be used to inexpensively test potential intervention strategies.